| Age | Commit message (Collapse) | Author |
|
For a long time now, this has done 3 different things, only one of which
is mypy. But they are labeled:
- LintMypy / lint
(runs pylint)
- LintMypy / custom_lint
(runs a custom script)
- LintMypy / mypy
(actually runs mypy)
This reduces the usable length of the label which isn't all that long to
begin with, is actively misleading, and even when you know what is going
on, it becomes harder to tell at a glance what failed. Change it to be
more unambiguous.
|
|
And the outdated pin means some error messages that are a lot better in
newer versions aren't available.
|
|
|
|
|
|
|
|
|
|
So we can lint it with mypy
|
|
Pin pylint version to workaround https://github.com/PyCQA/pylint/issues/3524
|
|
|
|
This is tested working with both msbuild and ninja/samu. Since our xcode
support is pretty much broken I didn't bother.
Fixes #6670
|
|
|
|
This also fixes that the keys in ArgumentNode.kwargs are
all of the type BaseNode now. Before this commit, it was
possible that both strings and Nodes where used as keys.
|
|
There were two things mypy was warning about:
1) it doesn't understand hasattr()
2) It was possible for mlog.{error,warning,deprecation} to get passed
multiple values for the once keyword argument.
|
|
|
|
|
|
|
|
This addresses various real-world problems with HDF5 pkg-config, including
* hdf*.pc with package versions as part of the filename
* malformed hdf*.pc missing the commonly-used HDF5 HL module
---
Additionally, this refactors more complicated dependencies such as
HDF5 and OpenMPI. This may help us deduplicate internal dependency
code in the future.
HDF5 selftest: improve platform-agnostic test
ci: init demo github action for HDF5 framework
ci Actions: hold off on MSYS2 for now [skip ci]
hdf5: ensure C libraries always included
ci: mac hdf5--use clang+gfortran
|
|
optimize intelMPI variables
mpi: extract version from non-OpenMPI wrapper
|
|
actions: run if .yml itself changed
|